The controller is designed by using the intelligent model of the TWIH which proposed in the previous work.
As mentioned before the TWIH system has no single mathematical model and the problem can be solved numerically for specific case.
The performance of the TWIH system tracking different desired temperature is shown in Figure (6) for aluminum workpiece of 15mm thickness and 25[degrees]C ambient temperature.
But, in the TWIH analysis there has not been applicable approaches to reduce computation time except single attempt to apply a method that combined a neural network with FEA to the design of a TWIH system to offset the resulting inhomogeneous eddy current or power density .
This paper proposed a novel method to estimate three main DOFs of the TWIH finite-element solution based on three stages neural networks.
The schematic configuration of typical TWIH system is shown in fig.
It can be seen from (7) that the current density in the strip is proportional to the rate of change of the magnetic flux and the field intensity, so magnetic field which is inducted by TWIH can create different vortex effect in the strip .